Brain Machine Interfaces are far from being used in games

Imagine playing your favourite game with the mere thought of your mind. That has been the dream of many sci-fi fans and has been seen in movies and books. In the last 10 years, research into brain machine interfaces (BMI) has provided some interesting studies.

For example, researchers in Netherlands at have managed to get two subjects to play a simple game of pong by reading their brain signals through fMRI.  Another group of researchers from the University of Pittsburgh and Carnegie Mellon managed to train a monkey to control a prothetic arm with his mind, and use it to grab and eat food.

So will we be mind fragging our opponents in Counterstrike and mentally micromanaging our zerg swarm in the near future?

No. There are many reasons these developments are a long way off, and one of the main reasons is the following: The methods used to read brain signals are either too inaccurate or too invasive.

Reading signals from the brain has been the staple for neuroscientists since the beginning of the field. However even with the many methods that we have to read the brain, each method has restrictions in terms of temporal and spatial resolution, invasiveness, and cost.

Electroencephalograms (EEGs) reads brain activity from electrodes placed on the scalp. EEGs are minimally invasive and have a good temporal resolution – meaning that there is very little delay between the measurment of the signal and the actual neural activity. The downside is that the signals have a very poor spatial resolution. The EEG signal reflects the activity of many millions of cells with very little specificity on where the signal is coming from. As such, using EEG signals would be very difficult, because understanding the meaning of the signals would be extremely hard.

Functional magnetic resonance imaging (fMRI) is another possible method. fMRI measures the changes in blood oxygen levels in the brain, which is associated with brain activity. As mentioned earlier, this method has been successfully used to play a simple game of pong. fMRI has a better spatial specificity than EEGs and is not invasive. However the downsides to this are that the fMRI signal has a low temporal resolution – the measured blood oxygenation response is occurs only 4-5 seconds after the corresponding brain activity. Four to five seconds may be trivial when you just need to move your paddle to hit a virtual ball, but if you want to snipe someone from across the map or engage your minions in battle you would need a better reaction time. Oh, I also should mention that the price tag for fMRI machines goes into the millions of dollars.

Finally there are single and multi-unit recordings. This method gets up close and personal to your neurons by inserting one or many electrodes directly into your brain. The advantages of these methods are they have both good temporal and spatial resolutions. The measurements can tell when and where your brain was active in the region that it is recording. In research, this method has been the most used. However, it is needless to say this method is extremely invasive. I am sure that you wouldn’t want to have your brain poked by a hundred little spikes for the sake of gaming.

So the non-invasive methods of brain recording (EEGs, fMRI) are not accurate enough, and the invasive methods (single and multi-unit recordings) are too dangerous. Actually, when I look at it, I think it is pretty ridiculous to even speculate the possibility of commercial brain machine interfaces.

Advertisements
This entry was posted in Uncategorized and tagged , , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s